Health Affairs, 25, no. 6 (2006): 1529-1536 doi: 10.1377/hlthaff.25.6.1529 © 2006 by Project HOPE	New Online   House Health Reform Bill   Paying for Reform   Vetting AHIP's Report   HIV/AIDS Costs   Brief: Insurance Reform   HA Blog Top 10

This Article Abstract Reprint (PDF) Submit a response to this article Alert me when this article is cited Alert me when Comments are posted Alert me if a correction is posted Services E-mail this article to a friend Similar articles in this journal Similar articles in Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to Citation Manager Reprints & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (6) Citing Articles via Google Scholar Google Scholar Articles by Busch, S. H. Articles by Cullen, M. R. Search for Related Content PubMed PubMed Citation Articles by Busch, S. H. Articles by Cullen, M. R. Related Collections Health Promotion/Disease Prevention Insurance Coverage Insurance - Employer-Based System Business Of Health Consumer Issues Insurance Market

Strategies

Effects Of A Cost-Sharing Exemption On Use Of Preventive Services At One Large Employer

Abstract

 
In 2004, Alcoa introduced a new health benefit for a portion of its workforce, which eliminated cost sharing for preventive care while increasing cost sharing for many other services. In this era of increased consumerism, Alcoa’s benefit redesign constituted an effort to reduce health care costs while preserving use of targeted services. Taking advantage of a unique natural experiment, we find that Alcoa was able to maintain rates of preventive service use. This evidence suggests that differential cost sharing can be used to preserve the use of critical health care services.

This paper evaluates recent efforts by Alcoa to provide incentives for employees and their families to maintain rates of preventive service use through differential consumer cost sharing. Alcoa is a U.S.-based, multinational producer of aluminum and related products with more than 40,000 U.S. employees in twenty-two states. Like other U.S. manufacturers, Alcoa is struggling with increasing health care costs that threaten to reduce its competitiveness in the global marketplace. The company introduced a new health benefit plan for a portion of its workforce in January 2004. This new benefit increased enrollees’ cost sharing for many outpatient and hospital services while simultaneously eliminating all cost sharing (both coinsurance and deductibles) for a wide range of preventive services. Alcoa’s benefit redesign was an effort to reduce overall costs but still maintain use of services deemed by the company to be important for employees’ health and to have the potential to reduce future health care costs.

To examine whether the benefit change affected the receipt of preventive services, we took advantage of a unique natural experiment. Because of existing multiyear union contracts, some employees were not subject to the benefit redesign. Using a pre-post study design, with employees not subject to the benefit change serving as a comparison group, we determined whether and to what extent Alcoa was able to maintain its enrollees’ preventive care use rates.

Background

Top Background Study Data And Methods Study Results Discussion NOTES

 
Cost-sharing trends. To reduce health care spending, many employers have increased employee cost sharing. Increased consumer cost sharing takes various forms, including higher deductibles, copayments, or coinsurance or the introduction of high-deductible health insurance products.³ The federal government encouraged this shift through enactment of the Medicare Prescription Drug, Improvement, and Modernization Act (MMA) of 2003, which grants beneficial tax treatment to health savings accounts (HSAs) associated with high-deductible plans. President George W. Bush recently proposed to expand the scope of HSA-qualified plans. The trend toward higher consumer cost sharing shows no sign of ebbing.

Concerns about cost sharing. One concern about increased cost sharing is that it might lead to declines in the use of services that have been shown to cost-effectively reduce the burden of disease, including preventive care. The RAND Health Insurance Experiment (HIE) provides the best evidence that people use fewer preventive services when they are required to pay part of the cost.⁴ Other research demonstrates that being uninsured (equivalent to 100 percent cost sharing) reduces rates of preventive service use.⁵ Yet other studies have found no change in preventive service use in response to small changes in cost sharing.⁶ Also, Geetesh Solanki and Helen Schauffler found that patient cost sharing had a negative effect on receipt of preventive care, with deductibles and coinsurance having a greater negative effect than fixed copayments per visit.⁷ In recognition of the potentially deleterious effects of cost sharing on preventive service use, the U.S. Congress included a safe-harbor provision in MMA permitting (but not requiring) firms to exempt preventive care services from the deductible in HSA-qualified plans.⁸

Health plan changes at Alcoa. In response to an annual U.S. health benefit burden of approximately $800 million, Alcoa has attempted to control spending growth while investing in a broad-based strategy to improve employees’ health and increase their productivity. Initiatives include restrictions on hours worked, introduction of new standards to reduce workplace injuries and occupational diseases, and encouraging employees to complete a health risk appraisal. Given the company’s emphasis on workforce health improvement, a decision was made to place special emphasis on preventive services under a benefit redesign implemented in 2004.

Prior to 2004, Alcoa offered a preferred provider organization (PPO)–type benefit package to employees and their dependents nationwide. For more than 98 percent of enrollees, the benefit had no deductible and a $10–$15 copayment for in-network visits.⁹ In January 2004, cost sharing for many services increased from $10–$15 a visit to a 10 percent coinsurance rate. Individual deductibles ranged from $0 to $1,500.¹⁰ Most preventive care services were exempt from both deductibles and coinsurance, eliminating consumers’ out-of-pocket cost for these services compared with the pre-2004 benefit.¹¹ Because of multiyear union contracts, the benefits of 53 percent of Alcoa’s hourly employees were exempt from these changes. The remainder of the hourly workforce and all salaried workers and the dependents of each group were covered under the new benefit. In this analysis we compare the preventive service use among hourly workers and their dependents who were subject to the new plan with that of workers who were not.

The study presented here builds on prior work by examining the demand response to lower cost sharing for preventive services in the context of an overall increase in consumer cost sharing for other services. Preliminary analysis indicates that the Alcoa benefit change resulted in lower total health care costs—a 5 percent reduction in total costs among affected enrollees, compared with a 4 percent rise in total costs in the comparison group. This spending reduction will be explored more extensively in future work.

Study Data And Methods

Top Background Study Data And Methods Study Results Discussion NOTES

 
Data source and study sample. As part of a unique collaboration between Alcoa and Yale University, we used outpatient claims data to examine the impact of benefit design changes on the use of preventive health care services. We compared differences in utilization rates among hourly workers and their dependents who were subject to the benefit design change (treatment group) with a similar group of hourly workers and their dependents who were not subject to the change (comparison group). To eliminate the possibility that plant-level differences explain our results, we also compared the effect of benefit changes for salaried workers at the same plants as our hourly comparison group. As noted above, all salaried employees and their dependents were subject to the new benefit in 2004. We limited our sample to those continuously enrolled for calendar years 2003 and 2004 and to Alcoa plants with more than 250 employees. By limiting our analysis to the continuously enrolled, we eliminated bias that could be introduced if the population was changing over time because of disenrollment. That we are looking at rates of use in the same individuals strengthens our confidence that any differences in the two years are attributable to the change in benefits.

Calculating utilization rates. Our outcome measures are preventive care utilization rates. Because men and women use these services differently, we calculated rates for men and women separately. We selected preventive services to study based on two criteria. First, the service had to be explicitly listed as exempt from copayments and deductibles in the benefit design booklet distributed to Alcoa employees. Second, we chose preventive services with strong clinical research indicating both effectiveness and cost-effectiveness. Services considered include cervical cancer screening (women ages 21–64), colorectal cancer screening (separately for men and women ages 50–64), well-child visits (all children ages 3–6), and adolescent well care (adolescents ages 12–21). As a robustness check, we also compared rates of breast cancer screening, since this service was exempt from cost sharing both before and after the benefit change.

We calculated utilization rates by examining whether the enrollee had a claim with a relevant diagnosis code, procedure code, or lab code for the service.¹² We followed the National Committee for Quality Assurance’s (NCQA’s) Health Plan Employer Data and Information Set (HEDIS) managed care measures as closely as possible. In some cases, the guideline indicates that a service should be received every two or three years. Since we have data for only one year before and one year after the benefit plan change, we could not assess whether an enrollee had received a service in a previous year. In some cases, there were exclusion criteria for a service that we were unable to assess using claims data (for example, a pap smear is indicated only for women with an intact uterus). Likewise, we were unable to determine conclusively whether a given procedure was preventive. For example, a person may present with a symptom and consequently be prescribed the relevant screening test. These caveats suggest that the rates presented here should not be interpreted as indicative of appropriateness of preventive care receipt in the Alcoa workforce generally. For the purpose of evaluating the effects of benefit design changes, by calculating utilization rates similarly in the pre and post periods, we minimized the likelihood that these discrepancies would bias results.

Study Results

Top Background Study Data And Methods Study Results Discussion NOTES

 
Characteristics of treatment versus comparison group. There are some differences in the characteristics of the treatment and comparison groups. As expected, among hourly workers and their dependents, men and women in the treatment group (which has fewer union members) are younger (thirty versus thirty-four years), have lower incomes ($32,967 versus $36,186), and have shorter tenures at Alcoa (thirteen versus nineteen years) (Exhibit 1). These differences made it critical that we compare changes within the same population over time.

Impact of plant location. We also considered the change in utilization rates for salaried workers at the same plants as hourly workers to determine if location affected changes in utilization. Salaried workers had somewhat higher rates of use of preventive service use than hourly workers. However, we found no evidence of any change in use (or differential change in use) among salaried workers at the plants with hourly workers in the treatment group versus salaried workers at plants with hourly workers in the comparison group (data not shown). This strengthens our confidence that our results are not attributable to either changes at the plants or secular changes in preventive service use.

Breast cancer screening. We next considered whether there were any changes in breast cancer screening rates, since this service was exempt from cost sharing in both the pre and post periods. We found no change in rates of use for this service.

Discussion

Top Background Study Data And Methods Study Results Discussion NOTES

 
With the aim of controlling costs, Alcoa initiated a change in health insurance coverage in 2004. We found that the benefit change had little impact on rates of preventive care use. Specifically, we found no change in rates of cervical cancer screening, colorectal cancer screening, or well-child/well-adolescent visits. Changes in preventive care use rates among those affected by the benefit change were essentially identical to changes in rates of use in a comparison group of Alcoa enrollees who were not subject to the new benefit. That a decline in copayment from $10–$15 to zero failed to increase preventive service use may be attributable to the relatively low initial copayment amount. This suggests that to increase use of these services, greater health education campaigns or even stronger financial incentives might be necessary.

Strengths of study design. Several strengths of our study design are worth noting. First, we took advantage of a natural experiment to analyze the impact of benefit changes in a real-world setting. The selection of those affected by the benefit change was determined at the plant level and therefore was exogenous to the specific characteristics of enrollees. As a result, our research design averted biases associated with self-selection into different coverage regimes. Since only a subset of hourly workers were subject to the new benefit, we were able to make comparisons both before and after the insurance change while simultaneously comparing those affected with an unaffected comparison group. Also, although this study was conducted in the context of a single manufacturer, the benefit change occurred across many U.S. locations, so our results might be broadly generalizable to other large employers that are considering similar benefit changes.

Study limitations. It is also important to note some key limitations to this analysis. First, since the new benefit was recently implemented, we could measure changes in utilization that occurred only in the initial year. A longer study period would be desirable to more precisely assess whether workers were receiving guideline-concordant preventive care. Also, enrollees’ behavior might change in the second year of implementation as they better understand the benefit changes. There are a number of potential weaknesses inherent in using outpatient claims data to measure receipt of preventive care services. This methodology relies on the accuracy of the codes recorded by providers. Moreover, preventive services can be delivered in venues not reported in an insurance claim. However, given the generosity of Alcoa’s preventive care coverage, we expect that few enrollees sought these services outside their insurance benefit.

Employers’ many choices. Although evidence suggests that high cost sharing results in substantial short-term cost savings to employers, use of low-cost services that reduce the burden of disease might also decline. This study suggests that differential cost sharing can be one partial solution to this dilemma.

In determining whether to exempt preventive care services from cost sharing, firms might differ in their broader objectives and priorities. Some firms might focus narrowly on cost control, while others might prioritize certain health benefits with the aim of increasing productivity. Still other firms might structure benefit design with an eye to attracting and retaining a highly skilled workforce. Some employers might be more interested in emphasizing services that lead to short-term productivity gains, rather than those such as preventive care with longer-term health impacts. Alcoa might have a greater incentive than other firms to invest in prevention because of its low turnover rate. The median tenure of Alcoa employees is 16.1 years, compared with the national median of 4 years.

Along with preventive services, there may be other services of value to employers or society that warrant greater protection from cost sharing. Mark Pauly and Philip Held contend that for interventions that are cost saving (that is, the expected value of future cost savings is greater than the cost of the intervention) but responsive to price, lower cost sharing may be optimal.¹³ Similarly, Alan Garber and others have argued that the value of a service should be considered when cost-sharing parameters are being set.¹⁴ Researchers have recently brought attention to the inherent conflict between increasing cost sharing while investing in disease management initiatives.¹⁵ In an era where many employers are contracting for the management of chronic disease, it might be sensible to exempt the health care services emphasized by these disease management programs from cost sharing.

That Alcoa increased cost sharing on most health care services and total costs greatly declined suggests that preventive care use rates might have fallen if the company had not decided to exempt prevention from cost sharing. In future work, it will be helpful to compare the results described in this paper with possible changes in utilization rates for acute and chronic care services that were subject to increased enrollee cost sharing.

Editor's Notes

 
Susan Busch (susan.busch{at}yale.edu) is an associate professor of public health at the Yale University School of Medicine in New Haven, Connecticut. Colleen Barry is an assistant professor of public health there; Sally Vegso is a biostatistician; and Jody Sindelar is a professor of public health. Mark Cullen is a professor of medicine and public health at the Yale School of Medicine and medical director at Alcoa Inc., also in New Haven.

The authors gratefully acknowledge support from the John D. and Catherine T. MacArthur Foundation Network on Socioeconomic Status and Health. Since 1997, the Yale Occupational Medicine Program has been the recipient of an ongoing service and research contract with Alcoa under which Yale faculty and staff perform jointly agreed-upon ongoing and ad hoc research projects on workers’ health, injury, disability, and health care. Under the agreement, Mark Cullen serves as medical director for the corporation.

NOTES

Top Background Study Data And Methods Study Results Discussion NOTES

 

1. D.P. Goldman et al., "Pharmacy Benefits and the Use of Drugs by the Chronically Ill," Journal of the American Medical Association 291, no. 19 (2004): 2344–2350.[Abstract/Free Full Text]
2. J. Gabel et al., "Health Benefits in 2005: Premium Increases Slow Down, Coverage Continues to Erode," Health Affairs 24, no. 5 (2005): 1273–1280.[Abstract/Free Full Text]
3. From 2004 to 2005, among covered workers facing copayments for physician visits in preferred provider organizations (PPOs), the share with copays of at least $20 increased from 47 percent to 58 percent. Among single workers, the share with a deductible of $1,000 or greater increased from 2 percent in 2000 to 9 percent in 2005. G. Claxton et al., Employer Health Benefits: 2005 Annual Survey, Tables 7.6 and 7.8, September 2005, http://www.kff.org/insurance/7315/index.cfm (accessed 18 August 2006).
4. J.P. Newhouse and the Insurance Experiment Group, Free for All? Lessons from the RAND Health Insurance Experiment (Cambridge, Mass.: Harvard University Press, 1994).
5. J.M. McWilliams et al., "Impact of Medicare Coverage on Basic Clinical Services for Previously Uninsured Adults," Journal of the American Medical Association 290, no. 6 (2003): 757–764[Abstract/Free Full Text]; and D.S. Kenkel, "The Demand for Preventive Medical Care," Applied Economics 26, no. 4 (1994): 313–325.
6. D.C. Cherkin, L. Grothaus, and E.H. Wagner, "The Effect of Office Visit Copayments on Preventive Care Services in an HMO," Inquiry 27, no. 1 (1990): 24–38[Web of Science][Medline]; and S. Tye et al., "Moving beyond the Typologies of Managed Care: The Example of Health Plan Predictors of Screening Mammography," Health Services Research 39, no. 1 (2004): 179–206.[CrossRef][Web of Science][Medline]
7. G. Solanki and H.H. Schauffler, "Cost-Sharing and the Utilization of Clinical Preventive Services," American Journal of Preventive Services 17, no. 2 (1999): 127–133.
8. U.S. Department of the Treasury, Internal Revenue Bulletin, "Notice 2004–23," Press Release, http://www.ustreas.gov/press/releases/reports/notice200423.pdf (accessed 12 September 2006).
9. Breast cancer screening and immunizations were not subject to enrollee cost sharing.
10. The distribution of deductibles was as follows: $0 (65 percent), $250 (26 percent), $500 (2 percent), $750 (1 percent), and $1,500 (6 percent).
11. The benefits literature specifically listed the following preventive health care as exempt from cost sharing: routine physical at any age, routine gynecological exam and pap smear, mammograms (no age-related limits), immunizations, well-baby care, and routine hearing screening for preventive purposes. Routine physical includes medical history, physical exam, prostate exam, proctosigmoidoscopic exam, pelvic exam, breast exam, blood pressure, total cholesterol, urinalysis, blood glucose, and electrocardiogram.
12. To determine diagnosis claims, we considered codes explicitly named in HEDIS and codes determined by the administrative-services-only (ASO) firms as indicating preventive care.
13. M.V. Pauly and P.J. Held, "Benign Moral Hazard and the Cost-Effectiveness Analysis of Insurance Coverage," Journal of Health Economics 9, no. 4 (1990): 447–461.[CrossRef][Web of Science][Medline]
14. A.M. Garber, "Cost-Effectiveness and Evidence Evaluation as Criteria for Coverage Policy," Health Affairs 23 (2004): w284–w296 (published online 19 May 2004; 10.1377/hlthaff.w4.284)[Abstract/Free Full Text]; and A.M. Fendrick and M.E. Chernew, "Value-based Insurance Design: A ‘Clinically Sensitive’ Approach to Preserve Quality of Care and Contain Costs," American Journal of Managed Care 12, no. 1 (2006): 18–20.[Web of Science][Medline]
15. M.E. Chernew, A.B. Rosen, and A.M. Fendrick, "Rising Out-of-Pocket Costs in Disease Management Programs," American Journal of Managed Care 12, no. 3 (2006): 150–154.[Web of Science][Medline]

                      What's this?

This article has been cited by other articles:

				C. L. Barry, M. R. Cullen, D. Galusha, M. D. Slade, and S. H. Busch Who Chooses A Consumer-Directed Health Plan? Health Aff., November 1, 2008; 27(6): 1671 - 1679. [Abstract] [Full Text] [PDF]

				A. P. Mahajan and R. H. Brook High-Deductible Health Plans and Better Benefit Design Ann Intern Med, May 6, 2008; 148(9): 704 - 706. [Full Text] [PDF]

				J. W. Rowe, T. Brown-Stevenson, R. L. Downey, and J. P. Newhouse The Effect Of Consumer-Directed Health Plans On The Use Of Preventive And Chronic Illness Services Health Aff., January 1, 2008; 27(1): 113 - 120. [Abstract] [Full Text] [PDF]